The present inventive concept herein relates to image sensors, and more particularly, to a depth camera for obtaining a three dimensional image and a method of synchronizing the depth camera.
As image processing technology develops, an interest in the field of modeling an object in a three-dimensional image is becoming great. Modeling an object in a three-dimensional image can be applied to a virtual reality movie and to a computer graphic of a video game. Thus, an image modeled in three dimensions is expected to be applied to various fields.
Three-dimensional image information includes geometry information and color information. The geometry information can be obtained using a depth image. The depth image can be directly obtained using filming equipment such as a depth camera. The depth image can also be indirectly obtained using an image process which is a computer vision technology, without using a depth camera.
In the method of obtaining the depth image using the depth camera, a method of measuring a time taken until an emitted light returns after it is reflected by an object is widely used. The time is referred to as “time of flight (ToF)”. To realize a three dimensional image in a more realistic angle, a multi-ToF method using a plurality of depth cameras is needed. To realize a three dimensional image using a plurality of depth cameras, synchronization of the plurality of depth cameras becomes an important issue. A trigger for initialization is periodically applied to a plurality of depth cameras to perform synchronization. However, in this method, a frame rate of the three-dimensional image is reduced and a deviation may occur in synchronization because of a distance difference between a host providing a trigger and the cameras. If using emitting lights having different frequencies or pulse widths, a depth error may increase due to interference between the depth cameras.
Thus, to obtain a three-dimensional image having high resolution and high performance in a multi-ToF method, a technology that can perform efficient synchronization of the depth cameras is needed.